The data consist of
scans of bullet land engraved
areas in x3p format. The
study is structured so
that there are 8 barrels
(identified with an alphanumeric code, such as A9, C8, etc.) used to
fire the known bullets in the
study, with three bullets
known to be fired from
each barrel (Bullet1, Bullet2, Bullet3).
Each bullet contains 6
land engraved areas, which are
identified as Land 1-Land
6.
In addition to the 24
bullets with known provenance,
there are 8 "unknown"
bullets which can be matched
to the 8 barrels, for
a total of 32 bullets.
Each bullet has 6 land
engraved areas, for a total
of 192 scans. Unknown bullets are identified with a letter.
This study is an open study, that is, not all of the unknown bullets may match to the known barrels in the set. Some of the unknown bullets may have been fired from other types of Ruger handguns.
The x3p format provides
for the inclusion of relevant
metadata within the x3p
file, including the machine used
to scan the bullets, the
operator, and the resolution
of the scan. The total dataset size is about 3.4 GB.
Files are grouped
as follows: BarrelXX/BulletY/ with the land number identified as part of the filename.
We have produced two R
packages meant to work
with this data:
- x3ptools (on CRAN, with
dev versions on github
at heike/x3ptools). x3ptools contains
utilities for reading, writing,
annotating, rendering, and exporting
x3p data to other formats.
- bulletxtrctr (dev version
on github at heike/bulletxtrctr).
bulletxtrctr contains utilities for
extracting useful data from
bullet LEA scans as well
as a matching algorithm
used to match unknown bullets
to the barrels they were
fired from.
Scans were created by Bill Henderson at Sensofar. Tylor Klep of the Phoenix Police Department designed the study and collected the bullets.
Funding
VPR Funds, 290-17-04-16-1000
National Institutes of Standards and Technology, 426-17-02